By ‘family structure’ many things may be intended. I shall take it here in two senses. First, in the sense of composition of the co-resident domestic group, as the historical sociologists call it. This means the knot of persons who live together, man, wife and sometimes, but by no means always, their children, their relatives, if any, along with their servants, now excessively rare. Such is the family which the wage-earner leaves when he catches his bus in the morning to go to work, and which he returns to in the evening. It is also the assemblage of possessions which the bachelor girl or the solitary widower or divorcee calls a home, along with himself or herself. A modest array of this kind is the constitution of the family for very large numbers of Western Europeans in the 1980s. The family in the second sense is the extended family of cousins, aunts, uncles and so on who are recognised, and sometimes associated with, but do not live together in the same place.
I am concerned here with family in both these senses, historically, over time. This is not only because of the interest people have in their family history: but also because it has now been shown that unless we have some knowledge of the history of the family, the family of today, of our own personal experience, can be profoundly misunderstood. For the fact is that in the matter of the family we have suffered, and still suffer, from a series of persistent, deceptive, obfuscating misbeliefs which can only be shown up and corrected by a knowledge of the past.
This self-deception about the history of the family has particularly affected Western Europeans. Frenchmen, Germans or Englishmen, unless they have come across the work of recent historical sociologists, are likely to believe the following. That the co-resident familial group in the past, at least up to the point of industrialisation, was large and complicated, with several generations living together. Furthermore, that this comfortable, kin-enfolding, welfare-providing family group not only nurtured the young, but took in their spouses when they married, and also provided them with shelter and succour when they became old or suffered other misfortunes. That the family in the sense of extended kin was a further source of welfare. It seems to be supposed that before the days of the Welfare State it was the family and kin which rescued social casualties. Now all this has turned out to be untrue.
Untrue, that is to say, in a literal sense and for the particular part of Western Europe which first became industrialised and which has given what might be called industrial culture to the rest of the world. By this North-West Europe, especially the British Isles, Northern France, the Low Countries and Scandinavia, is meant. In this cultural region family groups had been simple in composition and quite modest in size for many centuries before industrialisation. Married children only seldom lived with their parents, and two couples in one family household were quite unusual. It is true that the family group has become much smaller in the 20th century, servants have disappeared, and solitary living has grown enormously in our own day: but this did not happen during the process of industrialisation as ‘traditional society’ gave way to ‘modern society’ and cannot be called a transition to the ‘nuclear family’. The ‘nuclear family’ was there already.
The kin composition of the English family group was much as it is today in the 16th century, and had been so since the 1300s, the 1200s or even earlier, but with one very imortant structural difference. Servants lived in large numbers of families, and the presence of servants made the family groups of the rich large, and the family groups of the poor correspondingly small. In this area of the West, moreover, welfare never flowed along lines of kinship. The casualties of the system, the widows, the orphans, the poverty-stricken, were supported by the collectivity rather than the family.
The family structure I have just described was at its most homogeneous in England and the Low Countries, was less so in Germany and in Central Europe, and much less so in Southern Europe. France, indeed, seems to have been divided. Languedoc had larger and more complex households, especially in the mountains, than the rest of France. In spite of this important fact, the development of the Western European family can no longer be seen in the particular way which has been fashionable for the last generation, not even in areas where complicated family households have been common – areas like the Pyrenees, the Massif Central, or, which is very surprising, Tuscany and other parts of Italy.
It is a mistake to think that the transition in family matters which is now going on in the developing societies of the contemporary world repeats the earlier history of the Western European family. Nevertheless, if we take Europe as a whole, and not simply our Northern and Western area of it, we are faced with the recent discovery that in Russia things were very different. Among the serfs before they were given their freedom in the 1860s, the family was large and complicated, more so than for any other example of complicated families which has yet been found anywhere in the world, in the past or in the present. The fact is that, within the continent of historic Europe as a whole, family forms varied as widely as it is possible to do. This not withstanding the circumstance that in Britain and in other parts of North-Western Europe they were uniformly small and simple.
Misconceptions about family structure in Western Europe do not end with the points we have so far set down. It is widely believed, and here the grounds for the belief are almost as significant as the belief itself, that girls got married much younger in our pre-industrial past. This was so, it is supposed, for two ‘reasons’: first, the power the father had over all the women in his family, and second, the very large numbers of children born to every married couple. In order to have such a bevy of offspring a woman has to marry early. If people are questioned as to why they hold these views they almost always reply by citing evidence which has proved to be peculiarly exasperating to the historical sociologist. They refer to literature. Indeed, it is amazing how often the passage from Romeo and Juliet is quoted where Juliet’s mother says to Juliet, who was, we are specifically told, only 13 years of age:
Well think of marriage now: younger than you,
Here in Verona, ladies of esteem,
Are made already mothers. By my count
I was your mother much upon these years that you are now a maid.
Historical demographers have now shown that the audience which saw the first performance of Romeo and Juliet in the 1590s belonged to a population where for women the mean age at marriage was between 23 and 24, where only one woman in a thousand got married as early as 13, and where 85 per cent were 20 years old and more. What is more, these ages of marriage for women were early, not late, in comparison with what came after in the 17th and early 18th century in England. Among the bonded serfs of the Russian estates, marriage age was much lower, both for men and for women, if never as low as in Romeo and Juliet. For early marriage is the usual accompaniment of complicated kinship structure in the household.
Historical demographers have further established that large numbers of children were not the universal rule in the Western European households of the pre-industrial past. Over the three centuries of English history between the 1540s and the 1840s the number of births to a married woman was only about four. It has turned out in fact, that the mistake about the typicality of Juliet’s marriage is not simply a mistake about demographic history and about the nature of Western European family development. It is a mistake about the whole character of European culture and society.
It is not impossible that, at the time and in the place where the original story about Romeo and Juliet was composed – in 15th-century Italy, that is to say – a highly privileged girl like Juliet could have got married in her early teens or even that her mother had a child at that age. Marriage was much earlier there and then – it was, in fact, ‘non-Western’. But mother and daughter could only have behaved as the play implied on one condition: that they were both among the very small numbers of girls who matured sexually as much as two or three years before the average age, which at all times before the later 19th century was certainly well above 15 and probably over 16. We have at the moment no knowledge as to whether Shakespeare himself, or his audience, had any idea that what was being performed upon the stage was so very exceptional, so very unlike the experience of the audience.
There is, then, a twist in the relationship of cultural attitude and familial fact in the history of Europe. Russian literature also portrays a large-scale, kin-complicated family group, and for the Russian past this image seems to have corresponded to fact, at least in the serf areas of what is called Great Russia: in Western Europe it was false, whatever might have been the case in Languedoc or in Florence.
An odd definition of the West, you may think, which could exclude Tuscany and the city of Florence. But it is with a Western Europe so defined that we have to be concerned, with the Europe of Paris and of Lille, of London and of Amsterdam, of Copenhagen and of Frankfurt, the Europe whose familial model spread itself across the North American continent and is now the dominating norm for all developed and industrial society. I shall list the historically enduring characteristics of family structure in that cultural stream, with a reflection or two on how these characteristics may have affected the historical development of the West and of a world which has been so largely dominated by the West.
Characteristic 1 is late age at marriage for women. In the region described above, it has never been true for any community for any length of time that a majority of girls have married before the age of 20, although this is a near-universal rule in the rest of the world. There has, therefore, always been an interval between physical capacity to procreate and the beginning of procreation, a gap which is not found in the rest of humanity. Western Europeans have postponed marriage, avoided making use of their full procreative potential, avoided immature housewives who, if they were to look after their own families at all might well have to do so while still within the family group of their own parents.
Characteristic 2 is variability in marriage age, for women especially but also for men. Since marriage age was not fixed by cultural convention, family formation and the rate of growth of the population could, and in fact did, vary with economic opportunity and the availability of resources. It is easy to see how such a flexible relationship between population, the family and the means of subsistence might give Western Europe a productive economic advantage, and how in this sense its particular family structure could lie at the very heart of its enormous economic and technological success (for how else shall we characterise the industrial history of Europe over the last three or four centuries?). I ought, perhaps, to mention here that although no fixed age at marriage is characteristic of Europe, yet there must have been some convention which prevented almost every woman from marrying below a minimum age: Juliet was indeed entirely exceptional.
Characteristic 3 is a relatively high level of persons never marrying: at most times 5 or 10 per cent of the population, but it could also be in the region of 15 or 20 per cent. Marriage and household headship, and the full citizenship which only these could bring with them, was a goal to be sought in Western European history, open to every mature individual but not attained by all of them. You had to prove yourself, man and wife, responsible, capable and possessed of the means to run your own familial institution and support your children. The doctrines of Thomas Robert Malthus are certainly audible here. It was a vice to marry too early, or to marry at all, unless you could afford it, and misery awaited those who broke the rules.
The fourth characteristic is a brief age gap between spouses, an average of some two or three years in seniority between husband and wife over most of Western European history as compared with ten or 15 years in regions where the European familial pattern did not prevail. (In most places in the West and at most times, about a fifth of all wives were older than their husbands.) It might be going too far to call this characteristic of European family structure companionate marriage. But the phrase draws attention to the fact that marriage arranged by parents, where the dowry was crucial and the extended family so important, is not characteristic of marriage in European familial history as a whole but only of marriage among the aristocracy and the property-owners. Most marriages among ordinary people were made at the instigation of the parties themselves with parental connivance and consent. They were companionate unions in the sense that both the man and the woman had to bring experience, skill and resources for a reciprocal arrangement, and this meant that they were likely to be of roughly equal ages. It has been very different in China and Japan, in India, Pakistan and Africa.
The fifth characteristic is that kinship bonding has on the whole been lax in Western Europe. It is well-known that the kin systems of our cultural area have been bilateral, that in most Western languages kinship terminology is abbreviated, lacking all the richness of detail possessed by other linguistic systems, and that in modern Western industrial society kin relationships are relatively inconspicuous. There are occasions when relatives may be crucially important – in the crises of life, when a person moves, or gets a new job, or his spouse dies – but most of the time kin and neighbours play almost interchangeable roles in personal life, with neighbourliness and companionship at work on the whole more important than kinship. What is now being discovered is that this is no more a product of industrialisation and urbanisation than the small-scale nuclear family. The evidence is imperfect and difficult to study in England, but the indications are that kin relations were lax in this way as far back as the 17th, or even the 14th and 13th centuries. The salient fact is that support of a continuing kind, the support needed by those who were really impoverished as a result of widowhood, the death of parents or other misfortunes, was not the responsibility of the extended family and never flowed along the kin lines. As I have said, this was provided by the collectivity: the church, institutions of charity, the village community, the municipality and the nation as a whole. The welfare state, that quintessentially European institutional achievement, as we tend to think of it, is in this sense as old as European history.
Characteristic 6 is the existence of large numbers of life-cycle servants: younger men and women spending the period between puberty and marriage as resident employees and members of the family of other persons, usually more substantial than their parents. It is not surprising that service of this kind, service which is exploitative as well as supportive, which requires submission to patriarchal authority as well as to the imperatives of learning a skill, which imposes saving on would-be spouses as well as a very low standard of personal consumption in early life, should be characteristic of an individualist, exploitative yet accumulative society.
But the Western European institution of life-cycle service has its surprises and its puzzles too. An outstanding feature of these very large numbers of young people in service was that they moved from one household to another almost every year. The faithful and trusty retainer is one more of the literary myths about European family structure – myths in the sense of substituting the entirely exceptional circumstance for the wholly normal, as in the case of Juliet’s marriage. The examination of servants and service is one of the most difficult and complicated of all the undertakings of contemporary familial historians. And it is of great significance, too, since as a life-cycle stage which was also a queue for marriage, servitude fits very well into an integrated conception of European society, where family composition is symmetrical with economic organisation and general social structure. In one respect at least, common opinion about what happened in Western Europe during the process of industrialisation is correct: service did disappear with the coming of the factories and the eclipsing of rural life, and its past importance in the European social structure has had to be painfully recovered.
The seventh and final characteristic is the numerical predominance of the small-scale, simple or nuclear family household. In the area whose lineaments I have been sketching, Northern France, the British Isles and the Scandinavian countries, with many German areas as well, a proposition put forward in 1972 has so far resisted all attempts at refutation. It runs like this. Whenever a family is encountered in the relevant historical evidence it must be assumed, unless there is contrary proof, that the family was built upon the model of our own: a family of parents and children, a broken family of parent and child, or a person living alone. The contrary assumption, the responsibility largely of the great French sociologist Frédéric le Play, that it should be taken to be complex in character and exhibit what is called the stem family form, must now be regarded as definitively superseded.
This region of late marriage, of frustrated reproductive power, and, it must be thought, of frustrated sexuality as well, this region which gave birth to Puritanism and the doctrines of Sigmund Freud, has not been an area of relatively high childbirth outside marriage. Non-Western Europe, where marriage was early, and approved, habitual sexual expression therefore open to everyone over 20, had higher illegitimacy levels than Western Europe. In England at least, the higher the age at marriage the fewer the illegitimates.